基于集成概率模型的变阻抗机器人打磨力控制

doi:10.3785/j.issn.1008-973X.2023.12.002

浙江大学学报(工学版)

2023, Vol. 57

Issue (12): 2356-2366 DOI: 10.3785/j.issn.1008-973X.2023.12.002

机械工程、能源工程

基于集成概率模型的变阻抗机器人打磨力控制

郭万金1,2,3(

),赵伍端1,利乾辉1,赵立军2,4,曹雏清3,4

1. 长安大学道路施工技术与装备教育部重点实验室，陕西西安 710064
2. 哈尔滨工业大学机器人技术与系统国家重点实验室，黑龙江哈尔滨 150001
3. 芜湖哈特机器人产业技术研究院有限公司，博士后工作站，安徽芜湖 241007
4. 长三角哈特机器人产业技术研究院，安徽芜湖 241007

Ensemble probabilistic model based variable impedance for robotic grinding force control

Wan-jin GUO1,2,3(

),Wu-duan ZHAO1,Qian-hui LI1,Li-jun ZHAO2,4,Chu-qing CAO3,4

1. Key Laboratory of Road Construction Technology and Equipment, Ministry of Education, Chang’an University, Xi'an 710064, China
2. State Key Laboratory of Robotics and System, Harbin Institute of Technology, Harbin 150001, China
3. Post-Doctoral Research Center, Wuhu HIT Robot Technology Research Institute Limited Company, Wuhu 241007, China
4. Yangtze River Delta HIT Robot Technology Research Institute, Wuhu 241007, China

全文: PDF(2326 KB) HTML

摘要：

工业机器人对工件柔顺打磨作业的适应性差，为此设计机器人柔顺浮动力控末端执行器，基于集成贝叶斯神经网络模型的强化学习，提出主动自适应变阻抗的机器人打磨力控制方法. 所提方法根据打磨作业的接触环境信息，利用自助法获取小量数据的多次采样样本，训练集成贝叶斯神经网络模型以描述机器人打磨系统与工况环境交互作用，采用协方差矩阵自适应进化策略（CMA-ES）求解最优阻抗参数. 构建机器人打磨系统虚拟样机平台，开展叶片工件的打磨仿真实验，验证所提方法的有效性. 实验结果表明，所提方法在十几次训练后，能够将打磨力的绝对跟踪误差减小至较小值，较好地实现了机器人打磨系统的主动自适应变阻抗打磨力控制，提高了机器人打磨力控制的柔顺性和鲁棒性.

关键词： 工业机器人; 打磨力控制; 自适应变阻抗; 强化学习; 集成贝叶斯神经网络

Abstract:

A compliant floating force-controlled end-effector was designed, in order to resolve the problem of poor adaptability of industrial robots for the compliant grinding of workpieces. A robotic grinding force control method with the active adaptive variable impedance was proposed, using the reinforcement-learning based on the ensemble Bayesian neural networks model. According to the contact environment information of the robotic grinding, the multiple sampling samples from the small amount of data were obtained by the Bootstrapping method, and the ensemble Bayesian neural network model was trained to characterize the interactions between the robotic grinding system and the grinding condition environment. The optimal impedance parameters were solved by the covariance matrix adaptation evolution strategy (CMA-ES). A virtual prototype platform of the robotic grinding system was constructed. A robotic grinding simulation experiment of a blade workpiece was conducted, and the effectiveness of the proposed method was verified. Experimental results show that the proposed method reduces the absolute tracking error of the grinding force to a small value after a dozen training, realizes the active adaptive variable impedance for the grinding force control of the robotic grinding system, and improves the flexibility and the robustness of the robotic grinding force control.

Key words: industrial robot grinding force control adaptive variable impedance reinforcement-learning ensemble Bayesian neural network

收稿日期: 2023-03-14 出版日期: 2023-12-27

CLC:

TP 242.2

基金资助: 国家自然科学基金资助项目(52275005)；中央高校基本科研业务费专项资金资助项目(300102253201)；安徽省博士后研究人员科研活动经费资助项目(2023B675)；中国博士后科学基金资助项目(2022M722435)；哈尔滨工业大学机器人技术与系统国家重点实验室开放研究项目(SKLRS-2020-KF-08)；安徽省教育厅科学研究重点项目(KJ2020A0364)；高校优秀青年人才支持计划项目(2019YQQ023)

作者简介: 郭万金（1983—），男，副教授，博导，从事工业机器人打磨技术与主动柔顺控制研究. orcid.org/0000-0001-9654-0113.E-mail： guowanjin@chd.edu.cn

	服务
	把本文推荐给朋友
	加入引用管理器
	E-mail Alert
	作者相关文章
	郭万金
	赵伍端
	利乾辉
	赵立军
	曹雏清

引用本文:

郭万金,赵伍端,利乾辉,赵立军,曹雏清. 基于集成概率模型的变阻抗机器人打磨力控制[J]. 浙江大学学报(工学版), 2023, 57(12): 2356-2366.

Wan-jin GUO,Wu-duan ZHAO,Qian-hui LI,Li-jun ZHAO,Chu-qing CAO. Ensemble probabilistic model based variable impedance for robotic grinding force control. Journal of ZheJiang University (Engineering Science), 2023, 57(12): 2356-2366.

链接本文:

https://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2023.12.002 或 https://www.zjujournals.com/eng/CN/Y2023/V57/I12/2356

图 1 柔顺浮动力控末端执行器[14]

图 2 机器人打磨系统虚拟样机

图 3 变阻抗控制的控制框图

图 4 集成概率模型的训练过程

图 5 基于集成概率模型的主动自适应变阻抗的机器人打磨力控制框图

图 6 叶片工件、规划路径点和曲面法向量

图 7 机器人打磨的工具轨迹

图 8 机器人各关节的角度、角速度、角加速度及急动度曲线

图 9 不同打磨期望力在不同训练次数下的打磨力曲线

图 10 不同打磨期望力在不同训练次数下的奖励值曲线

表 1 打磨力绝对跟踪误差

图 11 不同刚度工件的打磨力曲线

表 2 不同刚度工件打磨力绝对跟踪误差

1	ZHU D, FENG X, XU X, et al Robotic grinding of complex components: a step towards efficient and intelligent machining–challenges, solutions, and applications[J]. Robotics and Computer-Integrated Manufacturing, 2020, 65: 101908 doi: 10.1016/j.rcim.2019.101908
2	黄云, 肖贵坚, 邹莱航空发动机叶片机器人精密砂带磨削研究现状及发展趋势[J]. 航空学报, 2019, 40 (3): 022508 HUANG Yun, XIAO Gui-jian, ZOU Lai Current situation and development trend of robot precise belt grinding for aero-engine blade[J]. Acta Aeronautica et Astronautica Sinica, 2019, 40 (3): 022508
3	LIU L, ULRICH B J, ELBESTAWI M A. Robotic grinding force regulation: design, implementation and benefits [C]// IEEE International Conference on Robotics and Automation. Cincinnati: IEEE, 1990: 258-265.
4	WANG Q, WANG W, ZHENG L, et al Force control-based vibration suppression in robotic grinding of large thin-wall shells[J]. Robotics and Computer-Integrated Manufacturing, 2021, 67: 102031 doi: 10.1016/j.rcim.2020.102031
5	LI D, YANG J, ZHAO H, et al Contact force plan and control of robotic grinding towards ensuring contour accuracy of curved surfaces[J]. International Journal of Mechanical Sciences, 2022, 227: 107449 doi: 10.1016/j.ijmecsci.2022.107449
6	ZHANG T, XIAO M, ZOU Y B, et al Robotic curved surface tracking with a neural network for angle identification and constant force control based on reinforcement learning[J]. International Journal of Precision Engineering and Manufacturing, 2020, 21: 869- 882 doi: 10.1007/s12541-020-00315-x
7	甘亚辉, 段晋军, 戴先中非结构环境下的机器人自适应变阻抗力跟踪控制方法[J]. 控制与决策, 2019, 34 (10): 2134- 2142 GAN Ya-hui, DUAN Jin-jun, DAI Xian-zhong Adaptive variable impedance control for robot force tracking in unstructured environment[J]. Control and Decision, 2019, 34 (10): 2134- 2142
8	李超, 张智, 夏桂华, 等基于强化学习的学习变阻抗控制[J]. 哈尔滨工程大学学报, 2019, 40 (2): 304- 311 LI Chao, ZHANG Zhi, XIA Gui-hua, et al Learning variable impedance control based on reinforcement learning[J]. Journal of Harbin Engineering University, 2019, 40 (2): 304- 311
9	ZHOU H, MA S, WANG G, et al A hybrid control strategy for grinding and polishing robot based on adaptive impedance control[J]. Advances in Mechanical Engineering, 2021, 13 (3): 1- 21
10	SHEN Y, LU Y, ZHUANG C A fuzzy-based impedance control for force tracking in unknown environment[J]. Journal of Mechanical Science and Technology, 2022, 36: 5231- 5242 doi: 10.1007/s12206-022-0936-6
11	ZHONG Y, WANG T, PU Y, et al An adaptive bilateral impedance control based on nonlinear disturbance observer for different flexible targets grasping[J]. Computers and Electrical Engineering, 2022, 103: 108388 doi: 10.1016/j.compeleceng.2022.108388
12	WU X, HUANG Z, WAN Y, et al A novel force-controlled spherical polishing tool combined with self-rotation and co-rotation motion[J]. IEEE Access, 2020, 8: 108191- 108200 doi: 10.1109/ACCESS.2020.2997968
13	MOHAMMAD A E K, HONG J, WANG D, et al Synergistic integrated design of an electrochemical mechanical polishing end-effector for robotic polishing applications[J]. Robotics and Computer-Integrated Manufacturing, 2019, 55: 65- 75 doi: 10.1016/j.rcim.2018.07.005
14	郭万金, 赵伍端, 于苏扬, 等无先验模型曲面的机器人打磨主动自适应在线轨迹预测方法[J]. 浙江大学学报: 工学版, 2023, 57 (8): 1655- 1666 GUO Wan-jin, ZHAO Wu-duan, YU Su-yang, et al Active adaptive online trajectory prediction for robotic grinding on surface without prior model[J]. Journal of Zhejiang University: Engineering Science, 2023, 57 (8): 1655- 1666
15	HOGAN N. Impedance control: an approach to manipulation [C]// 1984 American Control Conference. San Diego: IEEE, 1984: 304-313.
16	OTT C, MUKHERJEE R, NAKAMURA Y. Unified impedance and admittance control [C]// 2010 IEEE International Conference on Robotics and Automation. Anchorage: IEEE, 2010: 554-561.
17	肖蒙. 机器人打磨加工过程中恒力控制方法研究[D]. 广州: 华南理工大学, 2020. XIAO Meng. Research on constant force control methods in robot grinding process [D]. Guangzhou: South China University of Technology, 2020.
18	DEISENROTH M P, RASMUSSEN C E. PILCO: a model-based and data-efficient approach to policy search [C]// 28th International Conference on Machine Learning. Washington: Omnipress, 2011: 465-472.
19	张铁, 肖蒙, 邹焱飚, 等基于强化学习的机器人曲面恒力跟踪研究[J]. 浙江大学学报: 工学版, 2019, 53 (10): 1865- 1873 ZHANG Tie, XIAO Meng, ZOU Yan-biao, et al Research on robot constant force control of surface tracking based on reinforcement learning[J]. Journal of Zhejiang University: Engineering Science, 2019, 53 (10): 1865- 1873
20	KINGMA D P, SALIMANS T, WELLING M. Variational dropout and the local reparameterization trick [EB/OL]. (2015-12-20)[2023-03-12]. https://arxiv.org/pdf/1506.02557.pdf.
21	JOSPIN L V, LAGA H, BOUSSAID F, et al Hands-on bayesian neural networks: a tutorial for deep learning users[J]. IEEE Computational Intelligence Magazine, 2022, 17 (2): 29- 48 doi: 10.1109/MCI.2022.3155327
22	KINGMA D P, WELLING M. Auto-encoding variational bayes [EB/OL]. (2022-12-11)[2023-03-12]. https://arxiv.org/pdf/1312.6114.pdf.
23	CHUA K, CALANDRA R, MCALLISTER R, et al. Deep reinforcement learning in a handful of trials using probabilistic dynamics models [EB/OL]. (2018-11-02)[2023-03-12]. https://arxiv.org/pdf/1805.12114.pdf.
24	QUIRYNEN R, VUKOV M, ZANON M, et al Autogenerating microsecond solvers for nonlinear MPC: a tutorial using ACADO integrators[J]. Optimal Control Applications and Methods, 2015, 36 (5): 685- 704 doi: 10.1002/oca.2152
25	HANSEN N. The CMA evolution strategy: a comparing review [M]// LOZANO J A, LARRAÑAGA P, INZA I, et al. Towards a new evolutionary computation. Berlin: Springer, 2006, 192: 75-102.
26	普亚松, 史耀耀, 蔺小军, 等基于对数四元数的工业机器人Hermite样条曲线姿态插值[J]. 西北工业大学学报, 2019, 37 (6): 1165- 1173 PU Ya-song, SHI Yao-yao, LIN Xiao-jun, et al Interpolating industrial robot orientation with Hermite spline curve based on logarithmic quaternion[J]. Journal of Northwestern Polytechnical University, 2019, 37 (6): 1165- 1173 doi: 10.1051/jnwpu/20193761165
27	郭万金, 于苏扬, 赵伍端, 等机器人主动柔顺恒力打磨控制方法[J]. 东北大学学报: 自然科学版, 2023, 44 (1): 89- 99 GUO Wan-jin, YU Su-yang, ZHAO Wu-duan, et al Grinding control method of robotic active compliance constant-force[J]. Journal of Northeastern University: Natural Science, 2023, 44 (1): 89- 99

[1]	薛雅丽,叶金泽,李寒雁. 基于改进强化学习的多智能体追逃对抗[J]. 浙江大学学报(工学版), 2023, 57(8): 1479-1486.
[2]	郭万金,赵伍端,于苏扬,赵立军,曹雏清. 无先验模型曲面的机器人打磨主动自适应在线轨迹预测方法[J]. 浙江大学学报(工学版), 2023, 57(8): 1655-1666.
[3]	徐少铭,李钰,袁晴龙. 基于强化学习和3σ准则的组合剪枝方法[J]. 浙江大学学报(工学版), 2023, 57(3): 486-494.
[4]	姜玉峰,陈东生. 基于深度强化学习的大口径轴孔装配策略[J]. 浙江大学学报(工学版), 2023, 57(11): 2210-2216.
[5]	华夏,王新晴,芮挺,邵发明,王东. 视觉感知的无人机端到端目标跟踪控制技术[J]. 浙江大学学报(工学版), 2022, 56(7): 1464-1472.
[6]	刘智敏,叶宝林,朱耀东,姚青,吴维敏. 基于深度强化学习的交通信号控制方法[J]. 浙江大学学报(工学版), 2022, 56(6): 1249-1256.
[7]	徐小高,夏莹杰,朱思雨,邝砾. 基于强化学习的多路口可变车道协同控制方法[J]. 浙江大学学报(工学版), 2022, 56(5): 987-994, 1005.
[8]	李广龙,申德荣,聂铁铮,寇月. 数据库外基于多模型的学习式查询优化方法[J]. 浙江大学学报(工学版), 2022, 56(2): 288-296.
[9]	邓齐林,鲁娟,陈勇辉,冯健,廖小平,马俊燕. 基于深度强化学习的数控铣削加工参数优化方法[J]. 浙江大学学报(工学版), 2022, 56(11): 2145-2155.
[10]	马浩然,丁雅斌. 基于双目视觉的激光位移传感器标定方法[J]. 浙江大学学报(工学版), 2021, 55(9): 1634-1642.
[11]	马一凡,赵凡宇,王鑫,金仲和. 密集观测场景下的敏捷成像卫星任务规划方法[J]. 浙江大学学报(工学版), 2021, 55(6): 1215-1224.
[12]	马一凡,赵凡宇,王鑫,金仲和. 基于改进指针网络的卫星对地观测任务规划方法[J]. 浙江大学学报(工学版), 2021, 55(2): 395-401.
[13]	董大钊,徐冠华,高继良,徐月同,傅建中. 基于机器视觉的机器人装配位姿在线校正算法[J]. 浙江大学学报(工学版), 2021, 55(1): 145-152.
[14]	郭英杰,顾钒,董辉跃,汪海晋. 压脚压紧力作用下的机器人变形预测和补偿[J]. 浙江大学学报(工学版), 2020, 54(8): 1457-1465.
[15]	冯毅雄,李康杰,高一聪,郑浩. 面向视觉伺服的工业机器人轮廓曲线角点识别[J]. 浙江大学学报(工学版), 2020, 54(8): 1449-1456.

Viewed

Full text

Abstract

Cited

Shared

Discussed